How WindStream
Leverages EAL to Increase
their Analytic Capabilities
Alex Ladd
Sr. Partner
MindStream Analytics
Webinar will begin 12:03pm
MINDSTREAM
Agenda
• Introduction
• Audience Participation
• Windstream prior to EAL
• EAL Project Objectives
• The Solution
• DIA
• EAL
• Lessons Learned
• Questions
MINDSTREAM
Alex Ladd
• 15 Years of EPM Consulting Experience
• OAUG Hyperion SIG Essbase Domain Lead
• Senior Partner – MindStream Analytics
Contact Information
Email: [email protected]
Telephone: 917-579-0926
MINDSTREAM
Audience Participation
• How many people use HFM?
• How many people use HFM & Essbase (or Planning)?
• Do you have drill through?
o Do you want it?
• How many people use EPMA?
• How many of you want to keep EPMA?
MINDSTREAM
Windstream prior to EAL project
• Analytics desired to be done at higher level than HFM
• Analysis done in HFM then users switch to different tool
to launch queries of source data to back up analysis.
• Data loaded to HFM twice to show Pro-forma and Actual
• Management Reporting needs were difficult to handle in HFM
• Plan data done in HFM at some levels
MINDSTREAM
Windstream prior to EAL project - App
• 2 HFM Applications (Financial Reporting & Capital)
o ~20,000+ Accounts
o 13,000 Financial Accts & ~7,000 Stat Accts
o ~4,500+ Entities
o 3,500 Entities with 10 different rollup structures
o Many not used hierarchies in Custom
dimensions
MINDSTREAM
EAL Project Objectives
• Provide Better Analytic capabilities
• Provide Drillback to detail data
• Streamline HFM application
• Too many accounts and overly complex rules
• Make Management reporting easier
• Allow Planning to be done at a level higher than
HFM current level and allow drillback for FP&A
team to explain variances
MINDSTREAM
Initial Solution Design
SourceSystems
DataIntegra on
JDE
Applica onLayer
FinancialRepor ngandAnalysisTools
Hyperion-SmartViewHyperion-FinancialRepor ng
HyperionRepor ng&AnalysisEnvironment
HFM
Excel Uploads
(Statistical
Sources)
FDM
EssbaseEAL
Essbase
Drill ThroughDrill Through
MINDSTREAM
Design Discussions
• Essbase allows for faster aggregation of data. FDM/ERPi allows drill back into source systems to display transactions that make up the balances.
• In HFM, the base level contains over 10,000 Accounts, 496 FWGs (Cust1), 125 Product Values (Cust3) and over 3,200 Entities. That creates the opportunity for almost 2 trillion possible combinations. That is just for one Scenario. Multiply that by 5 for all the active scenarios.
• Drill back into JDE would require each drill back instance to be at a base level, i.e., each Entity, FWG, Account for this to work with the current HFM architecture.
• For this new tool to work efficiently and effectively for our users we determined that consolidation of Accounts and FWGs would be necessary.
MINDSTREAM
Unique Solution Points
• Drill back from multiple levels of the hierarchies• MindStream DIA solution to allow various Drill Capabilities
• Leverage JDE Category Codes
• Various applications connected but at different levels• Utilize Essbase as Data Integration Tool
• FDM pulling data from Essbase to Essbase or HFM
• Synchronization of Data between HFM & Essbase• Unique EAL Configuration
• Use of 2 Essbase databases to
MINDSTREAM
Final Solution Structure and Flow
MINDSTREAM
Design Explanations
• “Staging” is used to store the data at its lowest level and roll it up to a more manageable number of Accounts, FWGs, etc. Most users will not ever need access to this cube.
• “Reporting” will contain the financial data used in all reports. It contains the Entity structure as it is today, but the number of Accounts has been consolidated to just over 1,000 and FWG’s to 150.
• “Analysis” consolidates the Entity structure to Acquired/Heritage, with everything rolling into Heritage right now. FWGs are consolidated further to approximately 70. This closely matches the Cash Expense book and is intended for our accounting group’s use during close in order to drill back in JDE easily and get activity at this level.
• “Stats” contains non-financial statistical data as in Units/Customer Counts/Sales Productivity. It will share the same Entity structure as Reporting as well as the FWGs.
• “Planning” is owned by the Financial Planning group. It will share the same structure as Reporting.
• Product is no longer a dimension but is tagged on each entity and will no longer be a required field for loading. We have also added Legal Entity, Guarantor and Market information at the entity level.
MINDSTREAM
Data Extract Step 1 – JDE to Essbase
Data loaded to
Essbase via SQL Load
Rule
Essbase Staging
Aggregates data to
higher level than
previously in HFM
MINDSTREAM
Data Extract Step 2 – Staging to HFM
FDM Pulls data out of
Staging database and
feeds to HFM
DIA facilitates drill back
from FDM to JDE at
higher levels of detail
MINDSTREAM
Data Extract Step 3 – Staging to Reporting
FDM Pulls data out of
Staging database and
feeds to Essbase
Reporting database
DIA facilitates drill back
from FDM to JDE at
Account level 1 and base
FWG and Entity
MINDSTREAM
Data Extract Step 4 – Report to Analysis
FDM Pulls data out of
Reporting database and feeds
to Essbase Analytics
database which is at an even
higher level of detail
DIA facilitates drill back from
Analytics to JDE at Account
level 1, FWG level 2 and 1 of
three Entities
MINDSTREAM
Reporting and Analysis Solution
MINDSTREAM
Example of Staging Versus Reporting
If you wanted to find General Travelexpense recorded on BU 017000020in JDE:
HFM Entity = 00017000
Custom1 = 020 (OSP_SE) BU CC 29
Account = 701201001 (Travel) Account CC 23
Mapping now is done in Staging so you no longer see the base FWG but a roll up of several FWGs. Accounts are done the same way. Entity will rollup as it does today except in Analysis which will all be Heritage.
MINDSTREAM
Dimensions and Data and Cubes
Data is input at base-level members of stage.
400
900
1300
200
700
400
All Other Expenses
Travel
701201001
701205005
Training
514201001
The data in Stage is aggregated and loaded to the level we report in the reporting and analysis cubes.
Data is entered at the base level members of the dimensions in JDE and then we load them into the Staging Cube. The values for the parent-level members are aggregated from their children. These aggregated amounts are then loaded into Reporting and Analysis so that you see Travel in total.In the illustration, the data values from the base level members are rolled up to the parent level members, Travel and Training.These are then rolled up to their parent, All Other Expense, and so on.
MINDSTREAM
Drill Into Anything Solution
Custom MindStream solution to allow drill through from
FDM into other source systems beyond what ERPi allows.
Customized for Windstream to allow drill through at 0-
level of only certain dimensions
MINDSTREAM
Drill Through Example Step 1
User starts in Smart View and would like details of the
Travel expense
MINDSTREAM
Drill Through Step 2
Initial Drill through takes the user into FDM
MINDSTREAM
Drill Through Step 3
Secondary Drill takes the user to transactional detail via
DIA
MINDSTREAM
Grids now allow Drill Back into JDE for
transactional data by month
• In Reporting Cube – You can drill to the Base account and Entity and FWG. • Things to remember:
• Pro Forma is not base (Actual is base level)• ERPi is always JDE drill back. You are given an option of FDM and ERPi.
FDM would be there for anything that is not financial data• You always have to be a the base level to drill back and get results
• In Analysis Cube – Drill back is at the Account and FWG level but all data is loaded to one entity today. This entity is called Heritage. We have an entity called Acquired and will be loaded when we have a new acquisition so they can be reported separately.
MINDSTREAM
• EAL was originally proposed in order to allow HFM
and Essbase to stay in sync so that adjustments that
would affect pro forma reporting would be in sync
• Since the different dimension leveling in the overall
solution what was really needed was just a way to
move Journal entries in HFM to Essbase
EAL Used for?
MINDSTREAM
• EAL target database can’t be “owned” by HFM
• EAL was only moving Journals not all HFM data
• If EAL overwrote all Essbase Actual data the drill back
links would be lost
• EAL Replicated partitions would overwrite total
account value with just Journal entry instead of
adding to total value
• EAL does not send over #Missing for emptied out
values when using a replicated partition
EAL Challenges
MINDSTREAM
• Must use Transparent
partition
• Set up Extra
“Transparent” Database
in Essbase application
EAL Solution – Transparent Partition
MINDSTREAM
HFM Region (Extraction) Definition
MINDSTREAM
• Must use Transparent partition
• Set up Extra “Transparent” Database in Essbase
application
EAL Solution – Transparent Partition
MINDSTREAM
EAL Solution – Link DBs with XREF
• Scenario already dense
• Link Existing Reporting database with new “Transparent”
Database using XREF
MINDSTREAM
EAL Target Translation
MINDSTREAM
Lessons Learned
EPMA is a problem when using with Essbase
EAL is very different between versions
Still getting big updates
Need Relational Table Access to update Mapping tables
Replicated Partitions need work
EAL doesn’t like to share the Essbase Database it’s using